An important frontier in sociology is computational ethnography – the application of textual analysis, topic modelling, and related techniques to the data generated through ethnographic observation (e.g., field notes and interview transcripts). I got this idea when I saw a really great post-doc present a paper at ASA where historical materials were analyzed using topic modelling techniques, such as LDA.
Let me motivate this with a simple example. Let’s say I am a school ethnographer and I make a claim about how pupils perceive teachers. Typically, the ethnographer would offer an example from his or her field notes that illustrates the perceptions of the teacher. Then, someone would ask, “is this a typical observation?” and then the ethnographer would say, “yes, trust me.”
We no longer have to do that. Since ethnographers produce text, one can use topic models to map out themes or words that tend to appear in field notes and interview transcripts. Then, all block quotes from fields notes and transcripts can be compared to the entire corpus produced during field work. Not only would it attest to the commonality of a topic, but also how it is embedded in a larger network of discourse and meaning.
Cultural sociology, the future is here.
Subscribe to comments with RSS.
Comments are closed.